Thread Tools

**TonyD** · August 15th, 05:24 PM

Mineral Oil Submerged Computer

I'm sure some of you have seen oil-submerged PCs before, and maybe even this one. I think this might be the coolest custom PC job I've ever seen, and really, I don't often take much interest in these kind of things.

Your thoughts? It rules and you know it.

**XtremeAaron** · August 15th, 06:12 PM

Seen that exact page. I read it awhile back and I believe it said something about not good for long use, needs cool-down time? Kind of defeats the purpose imo, but cool as hell non-the-less.

***martix_agent*** · August 15th, 06:45 PM

it'll last forever so long as you run it through a cooling system(suck as..a scale sized nuclear cooling tower.(or their radiator idea..)

How does it not short everything out?

**TonyD** · August 15th, 07:09 PM

Quote:

Originally Posted by XtremeAaron

Seen that exact page. I read it awhile back and I believe it said something about not good for long use, needs cool-down time? Kind of defeats the purpose imo, but cool as hell non-the-less.

Read the 2 month update if you haven't already. What they found out is that it peaked at a very high temp (88°C), but this was after 12 hours of maxed-out usage—basically any normal user would never reach this temp. After it was set back to idle from it's max temp it took 8 hours to return to it's idle temp of 37°C.

But on the page's update they change some things around and dramatically improve the PC's cooling ability. By adding a radiator that runs the oil through it and cools it, they dropped the max temp. to 45°C instead of 88. This is a very good temperature for prolonged maximum-load usage, but it could be dropped even more with the addition of a few fans to the radiator, but they decide against it. They didn't do any other tests with the new radiator, such as idle temp. or return-to-idle time, but common sense says that both have been dramatically improved along with the max temp.

The components don't short out because the mineral oil has practically no electrical-conductive properties. Believe it or not, but water, in it's absolute purest form, will not conduct electricity, either. I know that we are led to believe—not entirely inaccurately—that liquid (water) + electricity = bad news, but there is actually a lot of chemistry involved that determines how the material interacts with the electricity.

It should be noted, however, that even though this is a pretty safe thing to do, you still run a number of risks in trying something like it; there are no guarantees and you are still prone to **** things up—just something to remember.

But with all this said, I think this might be something I'd consider trying someday (but probably with old parts as just a "Ha! Look at this!" kind of thing).

***martix_agent*** · August 15th, 07:28 PM

Quote:

Originally Posted by TonyD

Read the 2 month update if you haven't already. What they found out is that it peaked at a very high temp (88°C), but this was after 12 hours of maxed-out usage—basically any normal user would never reach this temp. After it was set back to idle from it's max temp it took 8 hours to return to it's idle temp of 37°C.

But on the page's update they change some things around and dramatically improve the PC's cooling ability. By adding a radiator that runs the oil through it and cools it, they dropped the max temp. to 45°C instead of 88. This is a very good temperature for prolonged maximum-load usage, but it could be dropped even more with the addition of a few fans to the radiator, but they decide against it. They didn't do any other tests with the new radiator, such as idle temp. or return-to-idle time, but common sense says that both have been dramatically improved along with the max temp.

The components don't short out because the mineral oil has practically no electrical-conductive properties. Believe it or not, but water, in it's absolute purest form, will not conduct electricity, either. I know that we are led to believe—not entirely inaccurately—that liquid (water) + electricity = bad news, but there is actually a lot of chemistry involved that determines how the material interacts with the electricity.

It should be noted, however, that even though this is a pretty safe thing to do, you still run a number of risks in trying something like it; there are no guarantees and you are still prone to **** things up—just something to remember.

But with all this said, I think this might be something I'd consider trying someday (but probably with old parts as just a "Ha! Look at this!" kind of thing).

oh, when I hear mineral water I think Paint thinner sort of stuff..why mineral water and not distilled water I wonder?

**TonyD** · August 15th, 07:58 PM

It's mineral oil, not mineral water. From other websites I've found with oil-cooled PCs, distilled water has been tested and it doesn't work as well as they originally hoped or expected. One site tried filling up their case with distilled water, and it worked for a little while, but short circuits occurred after about five minutes of running. On following attempts the issue happened over and over.

There are many factors for this, but I don't know them all. One of the explanations that I came across was the idea that the water would, even though it was "pure," oxidize with some of the metals, creating new ions within the liquid and thus destroying the non-conductive property of the water. This is why we think of water with electricity as such a dangerous thing; it's really hard to come across water that is 100% pure, and even if it is, there are no guarantees that it will stay that way.

**CGras90** · August 15th, 08:19 PM

I thought the oil would screw up the fan on the power supply, but i guess it just kept spinning through the oil. Eventually I bet it would burn out from all of the pressure but its not like you need the fan with oil cooling..

**XtremeAaron** · August 15th, 08:47 PM

Quote:

Originally Posted by TonyD

Read the 2 month update if you haven't already.

Ah, no update when I saw it. Thanks for the info.

**TonyD** · August 15th, 09:59 PM

Quote:

Originally Posted by CGras90

I thought the oil would screw up the fan on the power supply, but i guess it just kept spinning through the oil. Eventually I bet it would burn out from all of the pressure but its not like you need the fan with oil cooling..

It says something about this. The increased friction for the fans probably would indeed wear them out quicker, but that's easily solved simply by taking the fans out. They say they left them in for two reasons: (1) it looks cool and (2) they figured it'd move the oil around a little bit, helping keep the temperature consistent throughout. If I were going to do this, though, I would probably just take all of the fans out.

Alpha · August 16th, 01:45 AM

I'd rather build a CO2 cooled system. I almost did it once, but gave up after I couldn't find a solenoid valve that could handle more then 100PSI (and if you regulate it it loses a lot of its cooling properties).

Basically the idea was to run liquid CO2 through a liquid cooling water block, and bleed it through a radiator with a fan pulling air inwards.

The CO2 would expand in the water block, rapidly cooling its copper surface contact. IT would also cool the radiator, and by pulling cool air in with a fan, you are cooling the whole system.

One problem- condensation.

On the water thing- I'm not sure about oxidization of metals inside chips and solders, but I think dust will dissolve in water and become an electrolyte.

***MedicDVG*** · August 16th, 01:50 AM

They have been using oil in power transformers for years for this very reason.

Fascinating read.. I too am one of those skeptics.. but I am normally that way. Looks like fun to build one though.. who knows.. maybe in my off time when I get back.. lol

**L1nUx_r0ot** · August 16th, 09:26 AM

You know, this actually sounds like something i would consider doing. Not for the benefit of oil cooling, but for the cool factor. I mean, I go to a computer school, i gotta show some guys up.

**TonyD** · August 16th, 11:37 AM

Quote:

Originally Posted by Alpha

On the water thing- I'm not sure about oxidization of metals inside chips and solders, but I think dust will dissolve in water and become an electrolyte.

This is probably very likely. I think we can safely say that distilled water is not the medium of choice for this kind of project.

Quote:

Originally Posted by L1nUx_r0ot

You know, this actually sounds like something i would consider doing. Not for the benefit of oil cooling, but for the cool factor. I mean, I go to a computer school, i gotta show some guys up.

You should do it... someone on the forum should do it, and it sounds like you have a better reason than anyone right now, haha.

One of the interesting things I learned while reading up on a few other builds of this sort is that "best" kind of oil to use would actually be motor oil. Of course, it would look rather ugly and probably be even messier to clean up.

**imhungry** · August 16th, 11:58 AM

i saw this a while back, seems pretty cool but so not worth the trouble of building. id rather just have a conventional i can move around and not worry about pouring oil all over.

that thing has to be heavy as hell too

**L1nUx_r0ot** · August 16th, 02:17 PM

Quote:

Originally Posted by TonyD

One of the interesting things I learned while reading up on a few other builds of this sort is that "best" kind of oil to use would actually be motor oil. Of course, it would look rather ugly and probably be even messier to clean up.

So will Quaker State help keep my horses when I overclock?

and on the topic of doing it, I may be getting a spare computer from someone in the near future, no clue on the specs. All i have to do is help them transfer files from it to their new computer and I get to keep it. That would be perfect for a tester, because I'd much rather make a mistake on something I don't value.

Also, I noticed they had an external sata drive outside of the tank. I was curious about running an IDE (I assume that's what's in the spare computer) drive inside the oil.

**TonyD** · August 16th, 02:31 PM

One of the computers I came across did run an IDE, but I don't remember if it was in the oil or not. For the PC that I linked I believe the reasoning for leaving it outside of the oil was because they didn't want to make a mess if they were to detach the hard drive; basically, it's just easier.

***martix_agent*** · August 16th, 05:59 PM

I assume a thicker oil weight would cool better, but would also take more work to remove the heat from. I wonder if there's a way to find a happy medium